PHILADELPHIA – Concerns are being raised over a possible link between osteoarthritis and an increased risk of cardiovascular disease, with the signal strongest among those with walking disability.
A recent British population-based cohort study reported an excess of premature death from all causes among patients with knee or hip osteoarthritis (standardized mortality ratio, 1.55) as well as disease-specific increases that were particularly striking for cardiovascular disease (SMR, 1.71). Further, OA patients were twice as likely to die if they had a walking disability (BMJ 2011;342:d1165).
This comes on the heels of a systematic review involving seven studies that found moderate evidence of increased mortality in patients with OA (Clin. Exp. Rheumatol. 2008;26:S120-4).
However, a new, population-based study from the Netherlands found no link between OA and future cardiovascular disease (CVD) among adults aged 55 years and older. On the other hand, the presence of disability significantly predicted cardiovascular disease, independent of other CVD risk factors and the presence of OA, Sita Bierma-Zeinstra, Ph.D., reported.
She suggested that OA-related disability and co-morbidity may explain the earlier findings, but many in the audience were not convinced.
Fellow presenter Dr. Peter Jüni, professor of clinical epidemiology at the University of Bern, Switzerland where, the British analysis was performed, said what’s most impressive about their data is that the annual cardiovascular mortality rate of 2% in patients with OA and walking disability mirrors mortality rates observed in patients with an established diagnosis of coronary artery disease.
"We’re talking about a frequency of cardiovascular deaths that is comparable with what you see in cardiological patients who actually deserve invasive treatment," he said at the World Congress on Osteoarthritis, sponsored by Osteoarthritis Research Society International.
Dr. Jüni said randomized trials are needed to better define the association between OA and CVD, but the bottom line is that clinicians need to start looking differently at OA as a prognostic marker and consider including it in their risk assessment and management strategy.
These data need to be brought to the attention of the OA community so clinicians can "monitor the cardiovascular status of these patients because they are at risk, and I don’t think that’s commonly known yet. There is a signal here that we need to pay attention to," commented Dr. Stefan Lohmander of Lund (Sweden) University and coauthor of the recent European League Against Rheumatism (EULAR) recommendations for the nonpharmacological core management of knee and hip OA (Ann. Rheum. Dis. 2013 [doi:10.1136/annrheumdis-2012-202745]).
"You are my hero; I love this work, it’s incredible," attendee Dr. Gillian Hawker, professor of medicine and rheumatology at the University of Toronto, rose to say following Dr. Jüni’s lecture.
She noted that her group has been following a very similar cohort in the Ontario Osteoarthritis study, and that unpublished data confirm the British findings of increased all-cause mortality and CVD mortality, particularly with increasing disability, when assessed using the Western Ontario and McMaster Universities Arthritis Index (WOMAC) disability subscale, but not its pain subscale.
"People think of it [OA] as normal aging, wear and tear, grin and bear it, but if we can make links to mortality or worsening of other conditions like cardiovascular disease and diabetes, which everyone cares about, then maybe we’ll start paying attention to disability from OA," Dr. Hawker said in an interview. "If we can get people to care, then maybe we can get some effective drugs."
One possible explanation for the conflicting results between the two studies is that the British study included patients with a different OA phenotype, including patients with hip or knee OA and Kellgren-Lawrence grades 1-3, whereas the Dutch study also included patients with hand OA and only K-L grades 2-3, Dr. Jüni suggested. The Dutch study also used a different, "considerably softer" CVD composite outcome, he said in an interview.
"The way our study was set up, we couldn’t look at things the same way that they did, but what came out is that risks are particularly pronounced in people with disability and this message doesn’t change," he added.
The Dutch investigators used radiographs and symptoms of knee, hip, and hand OA to classify 4,648 persons in the 1990-1992 cohort of the prospective Rotterdam Study as having any disability versus none. At baseline, patients were free of CVD, 61% were women and their mean age was 67.6 years. Radiographic evidence of knee OA was present in 21%, hip OA in 10%, and hand OA in 30%. Clinical symptoms of OA were seen in 7%, 3%, and 7%, respectively.